基于GM(1,1)的CSTR故障预报

应用灰预测模型对连续搅拌式反应釜进行了状态预报检验,并给出了的灰建模过程。结果表明,根据自身数据建立动态微分方程再预报自身发展的灰预测是可行的,该预测模型计算简便、结果可靠。该方法能够及早判明故障的发展趋势及对设备的危害程度,避免了灾难性的事故发生,为故障状态下的设备停车与否提供决策指导,维护设备安全和长周期的运行。

Fault Forecast of CSTR Based on GM (1,1)

The principle and method of the prediction model using GM (1,1 )for prediction condition on the Continuous Stirred-Tank Reactor (CSTR) are discussed.The field example examination and the modeling process are presented.The results show that the forecast model is easy and reliable. Using this method, one can make sure of the fault developing tendency and judge dangers to the equipment, thus helps to make decision if the equipment should be shut down. This method not only avoids catastrophic accident, but assures the equipment have a safe and long operation.